Enum: DataProcessType

Data-handling processes enumerated in Clause 5.10 and Clause 3.2 (data acquisition, annotation, preparation, quality checking, sampling, augmentation, drift and poisoning handling, etc.).

URI: iso22989:DataProcessType

Permissible Values

Value	Meaning	Description
data_acquisition	None	Collecting data from one or more sources
exploratory_data_analysis	None	Initial profiling of a dataset to understand its characteristics (Clause 3
data_annotation	None	Adding labels or other metadata to data items (Clause 3
data_labeling	None	Assigning target labels to records for supervised learning
data_preparation	None	Transforming raw data into a form suitable for analysis or training
data_cleaning	None	Detecting and correcting errors and inconsistencies in data
filtering	None	Removing data items that do not match selection criteria
normalisation	None	Rescaling features to a common range or distribution
de_identification	None	Removing or transforming personally identifiable information
data_quality_checking	None	Assessing completeness, accuracy, representativeness and bias of data (Clause...
data_sampling	None	Selecting a subset of records from a larger population (Clause 3
data_augmentation	None	Creating additional training examples via transformation of existing data (Cl...
feature_engineering	None	Constructing or selecting features used as model inputs
imputation	None	Replacing missing values with substituted estimates (Clause 3
data_drift_detection	None	Detecting changes in the statistical distribution of operational data
data_poisoning_detection	None	Identifying adversarial contamination of training data
concept_drift_handling	None	Detecting and responding to changes in the relationship between inputs and ta...
catastrophic_forgetting_mitigation	None	Strategies to prevent loss of previously learned knowledge during retraining ...
retraining	None	Updating an existing trained model on new or revised data (Clause 5

Slots

Name	Description
process_type	Type of data-handling process performed

In Subsets

• Terminology

Identifier and Mapping Information

Schema Source

• from schema: https://w3id.org/lmodel/iso22989

LinkML Source

name: DataProcessType
description: Data-handling processes enumerated in Clause 5.10 and Clause 3.2 (data
  acquisition, annotation, preparation, quality checking, sampling, augmentation,
  drift and poisoning handling, etc.).
in_subset:
- terminology
from_schema: https://w3id.org/lmodel/iso22989
rank: 1000
permissible_values:
  data_acquisition:
    text: data_acquisition
    description: Collecting data from one or more sources.
  exploratory_data_analysis:
    text: exploratory_data_analysis
    description: Initial profiling of a dataset to understand its characteristics
      (Clause 3.2.6).
  data_annotation:
    text: data_annotation
    description: Adding labels or other metadata to data items (Clause 3.2.1).
  data_labeling:
    text: data_labeling
    description: Assigning target labels to records for supervised learning.
  data_preparation:
    text: data_preparation
    description: Transforming raw data into a form suitable for analysis or training.
  data_cleaning:
    text: data_cleaning
    description: Detecting and correcting errors and inconsistencies in data.
  filtering:
    text: filtering
    description: Removing data items that do not match selection criteria.
  normalisation:
    text: normalisation
    description: Rescaling features to a common range or distribution.
  de_identification:
    text: de_identification
    description: Removing or transforming personally identifiable information.
    close_mappings:
    - iso29100:PIIProcessingOperation
  data_quality_checking:
    text: data_quality_checking
    description: Assessing completeness, accuracy, representativeness and bias of
      data (Clause 3.2.2).
  data_sampling:
    text: data_sampling
    description: Selecting a subset of records from a larger population (Clause 3.2.4).
  data_augmentation:
    text: data_augmentation
    description: Creating additional training examples via transformation of existing
      data (Clause 3.2.3).
  feature_engineering:
    text: feature_engineering
    description: Constructing or selecting features used as model inputs.
  imputation:
    text: imputation
    description: Replacing missing values with substituted estimates (Clause 3.2.8).
  data_drift_detection:
    text: data_drift_detection
    description: Detecting changes in the statistical distribution of operational
      data.
  data_poisoning_detection:
    text: data_poisoning_detection
    description: Identifying adversarial contamination of training data.
  concept_drift_handling:
    text: concept_drift_handling
    description: Detecting and responding to changes in the relationship between inputs
      and target labels (Clause 5.11.9.1).
  catastrophic_forgetting_mitigation:
    text: catastrophic_forgetting_mitigation
    description: Strategies to prevent loss of previously learned knowledge during
      retraining (Clause 5.11.9.1).
  retraining:
    text: retraining
    description: Updating an existing trained model on new or revised data (Clause
      5.11.9).